Economic statistics at the end of the decade

Encontro Nacional
de Produtores e Usuários
de Informações Sociais,
Econômicas e Territoriais

INFORMAÇÃO PARA UMA SOCIEDADE MAIS JUSTA

III Conferência Nacional IV Conferência Nacional

de Geografia e Cartografia de Estatística

Reunião de Instituições Produtoras
Fórum de Usuários
Seminário "Desafios para Repensar o Trabalho"
Sitripósio de Inovações
Jornada de Cursos
Mostra de Tecnologias de Informação

27 a 31 de maio de 1 996
Rio de Janeiro, RJ BRASIL

Uma dos maneiras de alhar o ofício de produzir
informações saciais, ecanamicas e territoriais é como arte de
descrever o mundo. Estatísticas e mapas transportam os fenômenos
da realidade para escalas apropriadas à perspectiva de nossa visão
humana e nos permitem pensar e agir d distância, construindo
avenidas de mão dupla que juntam o mundo e suas imagens. Maior o
poder de síntese dessas representações, combinando, com precisão,
elementos dispersos e heterogêneos do cotidiano, maior o nosso
conhecimento e a nossa capacidade de compreender e transformar o
realidade.

Visto como arte. o ofício de produzir essas informações
reflete a cultura de um País e de sua época, como essa cultura vê o
mundo e o torna visível, redefinindo o que vê e o que hd poro se ver.

No cenário de contínua inovação tecnológica e mudança
de culturas do socledode contemporânea, as novas tecnologias de
informação - reunindo computadores, telecomunicações e redes de
informação - aceleram aquele movimento de mobilização do mundo
real. Aumenta o velocidade do acumulação de informação e são
ampliados seus requisitos de atualização, formato - mais flexível,
personalizado e interativo - e, principalmente, de acessibilidade. A
plataforma digital vem se consolidando como o meio mais simples,
barato e poderoso poro tratar o informação, tornando possíveis
novos produtos e serviços e conquistando novos usuários.

Acreditamos ser o ambiente de conversa e controvérsia
e de troca entre as diferentes disciplinas, nas mesas redondas e
sessões temáticas das Conferências Nacionais de (3eografia,
Cartografia e Estatística e do Simpósio de Inovações, aquele que
melhor enseja o aprimoramento do consenso sobre os fenômenos o
serem mensurados poro retratar o socledode, o economia e o
Território nacional e sobre as prioridades e formatos das informações
necessórias poro o fortalecimento da cidadania, o definição de
políticas públicas e a gestão político - administrativa do País, e para
criar uma socledode mais Justa.

Simon Schwartzman
Coordenador Ceral do ENCONTRO

Promoção

Fundação Instituto Brasileiro de Geografia e Estatística

IBGE

Fundação Instituto Brasileiro de Geografia e Estatística

IBGE

Associação Brasileira de Estudos Populacionais

ABEP
Co-Promoção
Associação Brasileira de Estatística

ABE

Associação Brasileira de Estudos do Trabalho

ABET

Associação Brasileira de Pós-graduação em Saúde Coletiva

ABRASCO

Associação Nacional de Centros de Pós-graduação em Economia

ANPEC

Associação Nacional de Pós-graduação e Pesquisa em Ciências

Sociais

ANPOCS

Associação Nacional de Pós-graduação e Pesquisa em Geografia

ANPEGE

Associação Nacional de Pós-graduação e Pesquisa em

Planejamento Urbano e Regional

ANPUR

Sociedade Brasileira de Cartografia

SBC

Apoio

Federação das Indústrias do Estado do Rio de Janeiro

FIRJAN

Academia Brasileira de Letras

ABL

Conselho Nacional de Pesquisas

CNPq

Financiadora de Estudos n Projetos

FINEP

Movistu (jíMicia Hoje

Institutos Regionais Associados

Companhia do Desenvolvimento do Planalto Central
CODEPLAN (DF)

Empresa Metropolitana de Planejamento da Grande São Paulo S/A
EMPLASA (SP)

Empresa Municipal de Informática e Planejamento S/A
IPLANRIO (RJ)

Fundação Centro de Informações e Dados do Rio de Janeiro
CIDE (RJ)

Fundação de Economia e Estatística
FEE (RS)

Fundação de Planejamento Metropolitano e Regional
METROPLAN (RS)

Fundação Instituto de Planejamento do Ceará
IPLANCE (CE)

Fundação João Pinheiro
FJP (MG)

Fundação Joaquim Nabuco
FUNDAJ (PE)

Fundação Sistema Estadual de Análise de Dados
SEADE (SP)

Instituto Ambiental do Paraná
lAP (PR)

Instituto de Geociências Aplicadas
IGA (MG)

Instituto de Pesquisas Econômicas, Administrativas e Contábeis
IPEAD (MG)

Instituto do Desenvolvimento Econômico Social do Pará
IDESP (PA)

Instituto Geográfico e Cartográfico
IGC (SP)

Instituto de Apoio à Pesquisa e ao Desenvolvimento “Jones dos
Santos Neves”

IJSN (ES)

Instituto Paranaense de Desenvolvimento Econômico e Social
IPARDES (PR)

Processamento de Dados do Município de Belo Horizonte S/A
PRODABEL (MG)

Superintendência de Estudos Econômicos e Sociais da Bahia
SEI (BA)

3

Organização

Coordenação Geral

Simon Schwartzman

Confege

César Ajara (IBGE)

Denizar Blitzkow (USP)

Jorge Marques (UFRJ)

Lia Osório Machado (UFRJ)
Mauro Pereira de Mello (IBGE)
Speridiâo Faissol (UERJ)
Trento Natali Filho (IBGE)

Comissões de Programa
Confest

José A. M. de Carvalho (UFMG)
José Márcio Camargo (PUC)
Lenildo Fernandes Silva (IBGE)
Teresa Cristina N. Araújo (IBGE)
Vilmar Faria (CEBRAP)
Wilton Bussab (FGV)

Comissão Organizadora

Secretaria Executiva - Luisa Maria La Croix
Secretaria Geral - Luciana Kanham
Confege, Confest e Simpósio de Inovações

Anna Lucia Barreto de Freitas, Evangelina X.G. de Oliveira,
Jaime Frankiin Vidal Araújo, Lilibeth Cardozo R. Ferreira e

Maria Letícia Duarte Warner
Jornada de Cursos - Carmen Feijó
Finanças - Marise Maria Ferreira
Comunicação Social - Micheline Christophe e Carlos Vieira
Programação Visual - Aldo Victorio Filho e
Luiz Gonzaga C. dos Santos
Infra-Estrutura - Maria Helena Neves Pereira de Souza
Atendimento aos Participantes - Cristina Lins

Apoio

Andréa de Carvalho F. Rodrigues, Carlos Alberto dos Santos,
Delfim Teixeira, Evilmerodac D. da Silva, Gilberto Scheid,
Héctor O. Pravaz, Ivan P. Jordão Junior,
José Augusto dos Santos, Juiio da Silva, Katia V. Cavalcanti, Lecy Delfim,

Maria Helena de M. Castro, Regina T. Fonseca,
Rita de Cassia Ataualpa Silva o T.air.;i Sawc/uk
Registramos ainda a colaboração de técnicos das dileientes
áreas do IBGE, com seu trabalho, cólicas e sugestões p.ii.i a
consolidaç.io do pio|eln do I NUON I HO.

4

National Symposium of Produccrs and Users of Social, Econoniic and

Territorial Information
May 27-31, J996
Rio de Janeiro, Brazil

“Economic Statistics At The End Of the Decadc”

by

Jacob Hyten
Statistics Canada

1 dúnt know abuul lhe end of the decade cxcept thal Ihcrc is a problem with the Computer
Systems of my ínstitution. la fact this is the oiüy problem about the year 2000 of \diicb I
am sure. Whea tíie year 2000 arrives many of the programmes on which we rely today
will have lo be changed at gieat troubie and expensc bccause they depend on routines lhal
limil themselves to comparing the two last digits of the year and assume stupidly that
cach year is denoted by a numbta* b^ger than íhc number denoling its predecessor.

In the world outside our ttffícos, changes will nol bc as sharp. The cconomic and social
world will bc continuing its relenti(^ pace of change. But as it has happened In lhe past,
we shall be labouring tmdcr dic sness o^ited by Üfô inadeqtiacy ofinteilectual schemes
which were helpM in the p^ but are of incn^ingly scant valuc in the face of current
developments. If anydiing wc may be sufferíng from a widening gap causcd by
»:celerated change in the real world and slow adaptation at constant speed on the part of
the way in which we staüsücians do and think about tbings.

Talking about lhe year 2000 is but a Symbol of our eoUectivc fhistration at the inadequate
way in 'udiich we dea! with those en^is thal are bothering us today and which we think
are expressíons of a persi^nt trend. It is also a slogan in the campwgn that wc are bound
to wage in our cry to mobília resourees and seek attentioa to the problcms with which
we shall confiont our users in a fcw yeara lime, Fortunalely, if wc abstract from our pre-
budgetaiy postures and look at the ftrtrae coldly, what is about to hit us in the year 2000
cannot be all that different firom what wc have been pcrcciving with gradual claiity in the
last few years. We know from observation thal those trends thal arc getting sbaiper arc
also the most likely to hit us hardest

1 have seleclcd a few of lhe perccivcd changes in cconomic organization, the
measurranent of which poses questions for which we in Canada have no rcady answers. 1
believe those changes should concem you as much as they concem us becausc if you do
not experíence títem today with the same force as wc do, it will not take long beforc we
are on a par. Tlús will be parlicularly the casc as you integrale with other countrícs and
your pattems of industrial organization rcflect as they arc bound to the process of
integratioiL Tn fact, 1 am convinced that the halhnark of the decade ushcred in by the year

2

2000 is that no country or gronp of countries will be ablc to ínsulate itself from what is
happening clsewfaere.

In order to discuss thc changes thal 1 expcct wil! trouble us > ofHcial statisticians - most, I
bove organized my paper into iivc dístinct sections:

Should laws stop at biirders?

Aro cstablishments dcad?

Wagc camcrs or hired gnns?

Aro bytes more important (han dollars?

Aro wc more produetive than we look?

At the end of each question 1 ask - noí entircly rhetorically - whcthcr you in Brazü or you
in Mercosur as thc case may be, have rccognized in your systems of measurement, legal
arrangements, or conceptual fhuneworlcs the relevance of thc questions with which I
have titled my five sections. The questions are inlcndcd to give ríse lo discussion
particularly where you fcel that the problem is as interesting to you as it is to u$ in
Canada.

Should latvs stop at borders?

We have a law. it extends to our geo-political border; but modem polilies and modem
industrial organization have made nonsensc of thc notion of bonier. Lct me give you a
few examples. Wc have a highly integrated economy with thc United States. In another
decade, the wel^t of NAFTA will lead us to addilional integration with México. What
does integration mean? That our industrira operale on both sides of lhe border under tlie
same ownership and control; that our trade iinions bargain collectívdy with their
management on behalf of membership on both sides of thc border; and thal thc benefits of
rescaroh and devclopment on cither side of thc border aocrue to both sides for Ihose
industries imder lhe same ownership and control.

In order to govem wisely, our authoriücs require to know thc State of our produclion and
the rcsources that go into it as much as the benefits that acerue from productíon to labour,
management and to stockholders. But manifestíy, a complete picture of the State of
aiTairs can only be gathered if thc informatíon does not stop at the 49th parallel but rather
ai lhe point where enteiprlscs operating on both sides of the border have their boundaries.
This is as truc for us as for our ncighbours to lhe South and as it will be for México when
thc extern of integration rcaches the same extent as it has reached betwccn Canada and
thc United States. At onc time we thought wc wcre unique in Ihís requirement. But in
thc fiamework of lhe Europcan Union, individual countrie-s experience thc same need for
informatíon and will continue to do so, as long as their govcrnmente are ultímatcly
responsible for admínistering fiscal and monetary policies and claim some autonomy.
Thus France and Germany and Spain to name but three of lhe Union raembers have cither

3

cxpcfienced or will expcrícnoc the need to know more about their resident enterpriscs that
operate abroad and conversely somcthing about their non-resident enterprises that operalc
in tíieir own countries. 'llie statistied oilíccs of all three countries are very much aware
that such modem industries as tclc-comniunications and informaties are central lo thc
operation of all transnational enterprises and that thc activities of Ihose industries do not
stop at any geo>politica1 border.

Consider a relatcd cjoimple. Take a grotq) of countries in an economic bloc. NAFTA is
not yet an economic bhc but the Hurapean Union is one. Such countries will create or
have created politlcal instituiions which in tum will have a say on how advantages and
rcsponsíbilitíes will be shared within the hloc. Normally, tíicrc will be a sharíng formula
and in order to ensure that its objcclivily is maintained it will be driven by ofücial
statisties. Hcncc the need for statistical comparability and also for additivity • so that
each country can derive a stated proportion of the advantages and discharge in an cqually
statcd proportion lhe responsibilities. But additivity of GDP, for example, is only
possible if wc do not duplicate or miss any of its components. However» in diis modem
era of imporlant enterprisc hcadquarters which provido a considerable range of Services to
their branches beyond nationd boundaries, it is not possible to either add or distribute
without duplicadon unless care is takcn to identiíy die sources of duplication. This is
only possible if there is a supra-national authoríty to whom access to individual
information on businesses is granted - typicaliy in contravention of what is
unambiguously statcd in any country's law about statistical confidentiality.

Bui if this issuc is only recognixed wühin thc Europcan Union, it will not take long
before it becomes a problcm in NAFTA and in Mcrcosur if tiie two economic areas
prosper.

Have you in the context of Mercosur thought of ihese matters? Do yowr partners ’
taws allowyou to coUect buxiness acíivity information from non-residents? Doyour
laws allowyou to reciprocaíe information to non-residents? What powers do you
have ifinteresting” non-residents det^ providing you with relevant information?
These are the conundrums that we shall experience increasingly and ones that our
body politic is only dimly aware of.

Are establishments dead?

Not so long 1 ^ 0 1 was asked to contribute to a Symposium on business statisties a paper
on the Outlook for cstabli^iunent statisties in ten ycars time.' I thought about thc subject
and dccidcd that on the basís oi’my expcrícncc with establishment statisties over the lasl

IBusiness Survey» ín Ten Ycars Time: EstabHshmonts Surveys, John Wiley, J993

4

twenty years there wajs no long tcrm fotuff • for whal had bccn a constant ín Canadian (and
American) business statisties - thc cstabUshment as a statistical müt Today» if anything
my view has hardened.

My rcasoning was as follows: establishments were conccivcd of when economies such as
the Canadian operated mines and slecl mills» himber yards and físh canneríes. 'fhese
activities were localcd in fixed sites, empioyed spccialized (and xmionized) pcrsonncl;
produced well defined produets with equally well dcfíncd machínery. They were helped
by a slew of Services chief among (hem transportotion, distribution and to a certain extení
Communications but thc latter did not enjoy an independent existence. Accordingly, if
things went badly for Steel and wood, all thc Services which supportcd those two
industries suffered as much as thc principal aclivily. And bcyond the svq>porting Services,
(here was not much else.

It is that image of thc economy which we have perpetuated widi our industriai
classifícations and our statistical units. But is it widely applicablc to conditions in
Canada today? Almost certainly not. While it is truc that we still have manufacturing
and resource bascd industries similar to those which we operated at the tum of thc
century, those industries have come down substantially as a sharc of ODP and today do
not account for even as much as one fifth of the total. Moreover, dieir importance as
cmployment creators and particularly as employment crealors for the young is greatly
diminished. Even in conjunction with mining, water and electrícity and construction all
of which are of great importance to thc Canadian economy, the sector does not account
for much more than a third of GDP.

Our Services - tclccommunications, finonce, general Services to cnteiprises- have come to
play an increasingly higher proíile role but the form in wdiich tho.se scrviccs are organized
is very difíerent from what cbaracterÍ7«d induslry carlicr in the century. For one they are
gypsy líke. Their localíon is ncilher long term permanent nor can it be easily interpreted
as physical factors play a marginal role and only a specialized fraction of thc labour force
(wÜch in any ca.se is very mobile) is of interest. 71ic fírms in question have headquarters
at fixed addresses as well as placcs where equípment is repaired and so on. But the
relationship bctwccn those headquarters and maintcnance centres as well as the activities
in which they are engaged bear a ílimsy relationship to our traditional dcfmition of
establisliment.

Actually, profound changes have been affecting manufacturing both since the signing of
NAFTA and even more so since lhe onset of the last economic recession in thc carly
nineties. A niunber of companies have decided that they should operate their industrial
activities offshore and have lucaled them cithcr immediateiy South of thc Rio Grande or
else to lhe more dynamic Far Ea.stem countries to which so much of Western industry has
been increasingly moved.

5

There is no rcason lo suppose that this tiend is about to reverse itself. In lhe mcantimc,
we have to decide how and with the Help of which slalislical indicators, we are going lo
undersland what our enterprises do. Employmcnt data suggests unambiguously that
many of those enteiprises no longcr do manufacturing. We presume that they do
planning, designing, managing, fínancíng and selling and providing all lhe support
required for entities that operate (manufáclurc) abroad but we cannot asscmblc a
complete picturc of the enterprise if we do no more than delecling resident
eslablishments.

II follows tliat we should Ihink about other methods to carve up our rcspcctive universes
of business and pay more attention to the need to have a full measure of each busiitoss
that operates independenlly. This is the same as saying that we believe that the time is
ripe for a roovement away from the establishment and towards enterprise statisties for
only Ihcn will we realize where the distinel bits of the enterprise are localed and how dicy
reiate to each other.

When was the last time that you in Jirazil askedyourselves whether the Information
that you collect from establishments or local imits is the most qppropriate in view of
the wí^your indusiry is organized? And have you ever thought of designing a survey
that would answer the question whether you should stick to yow traditional tmiL^ or
you should move to anofher? What are your thoughts on the effects of a ftãlyfledged
Mercosur onyour industrial organization? And have you discussed this matíer with
your partners?

Wage earners and hircd guns

My daughter is an architcct. She works in an architecfs ofTice. It is a big office ' it
employs about onc hundred professionals induding architccts and engineers, town
planners, lawyers and so on. They are moslly young and rceently out of graduate school.
They have one feature in common - none of them is an cmployee, 'TTial is right ► they are
all self-employed. Al leasl ihis is what they report on their tax fomí and this is what their
employer reports on his. His is but an oílice with two or perhaps three employee.s - a
secretary, a receplionist and an aceountant All other labour is acquircd in the form of
purchased Services. The employees are eonsultants \dio sell advice to lhe firm.

I am not surc what is built in to the contracts of lhese young pcople. 1 presume they are
sufHciently hard-nosed lo insist on a provision for holidays» síck leave, regular revisions
to their contraets and general Insurance to cover illnc.s.s and disabílity. But of course
theyarc not insured against unemployment and if they want to build up privale pension
rights they have to do it out of their - not salary - but self-employed income.

T mentioned my daughtcfs firm to give some immediacy lo what appears to be a growing
trend in the modem world in many a firm engaged in the provision of technical Services.

6

The mârket in which such ürms operalc is profoundly afFected by cyclicai tums as wcll as
by merciless technologícal innovation. The last things employcrs wish to have on their
hands i$ a stafT of hlgh camíng bul unocctq)icd profcssionals who could only bc
dismisscd after some compensatíon.

But our traditional stalislics do not Icnd themselves easily to an interpretation of whal
happcns in my daughter^s firm. If wc look at it from the poinl of vicw of cstabüshment or
enterprise statistics, wc would find no more than a firm cmploying three people eaming a
disproportionately latge gross bnsincss íncome. Suppose the íirm worked at two different
places >> one housing management and the other housing all the profcssionals contractcd
by the firm. And suppose we idenlilied the lattcr as the produetive establishmcnt and the
former as an ancillaiy unit. The establlshment would truthfully report zero employccs -
unlcss it had an employed securíty guard - but list as íts outputs» plans and tcchnical
reports. Only the ancillary unit could report value udded as it would know the difFerence
betwccn its gross income and the value of its purchascd Services, Ilowever, it would
report a tiny value added compared to its gross revenuc and of course there would be no
way through industry statistícs (o modify lhe ratio by taking into accounl lhe value added
of all the individual architects and other profcssionals. And to which industry would wc
add them?

There is no reason why in the fmmework of our household survey we could get the kind
of Information that would help us desciibe what happcns in the architect*s firm for which
my daughtcr works - its occupalional distribution or a full measure of ils value added.
Grantcd thst at a very aggregaled value wc would gct cverything right - which in itsclf is
questionable - we would stíll bc Icft witfa imperfect measures on v^o works iíi
architccturc and how much they produce. Nor would our conventional measures of
employmettt and unemploymcnt deal adequately wnth free-lancc people who may work
one week and not the next out of dioicc rathcr dum necessity.

If you aiB in the same position as we in Canada have bccn for the last twcnty or m<u«
years you will have nolíccd lhe gap between establishment employmcnt and household
statistícs. Of coutsc you wíll know about multiple job holders but not necessarily aboul
people who have disiqjpeared from the payroll of firms but are technically spcaking fiill
time employc»s. There are incrcasingly more imaginative arrangemenls to manage the
labourmarket. There are the (irms that have no output but only have employccs. They
are the ones that " rent out*' thcír employccs to other lirms in the same group. There
arethc cottage industry arrangements - of which systems dcvclopment is a good examplé,
in which intellectual work is làrmcd out to people who stay at home - or abroad - and are
satisfaotorily but only remotcly conncctcd to their cmployer or customer to provide a
Service on a regular basis,

But much is missing from our measíurcancnl syslcm before we are entitlcd to fccl that we
can safely brief our govcmmcnts about what is happemng to our labour markels and
wdietíier we Imve been alarming them unduly or not enough. I would mention three

7

frame^KS that either need lo bc changcd or dsc need to be croated. The first onc is a
proper dassification of the status of labour inpuls. Tlicre are members of lhe labour force
^vho are neither employees nor consultants although they look like consultants on paper
(tax forms) and like employees in practic». Unrortunalcly, our dassification schemes do
not yet do them justice. The Voorburg Group on Services Statisties has volunteered lo
conducl an initial inquiry into the matter of lhe evolving status of employment and
oceupations and to have reporls discussed by lhe Autumn of 1996. Bul T am not sure that
by itself it can muster lhe necassary rcsourccs to reflect all that is new in labour market
organizalion.

Sccondly, we need some form of Iracking of pcople as they join lhe labour market and
decide lo opt for one status or another, We know somcthing about their opiing out of the
Labour Force and we know what brings them baek. We also know that once in the
Labour Force they are either at work» have part time work or at work full time. But that
is a very simple State of alTairs» which neither deals wilh the chronic ailment of under-
employment which is so much more difElcuit to defíne and to measure nor with the
modem forms of atlachmcnt bctween workers and the firms for which they work.

Anyhow, we are £ar fram a situatíon in which wc can expiam why peoplc like my
daughter have gone fbr lhe slatus of hired gun while oüier professionals have prefeired
waiting but insisted on acquiríng lhe slatus of employees. We in Canada arc in Úic midst
of carrying out the first round of a very ambitious longitudinal survey of labour market
conditions. I bdieve that our model or improved version thcrcof wiU be adopted by
many more countries before lhe decade is over for rcasons similar lo the ones given
above.

Lastly, we need far botter dctcction mcthods than we have right now. Not long ago, I was
approached by soraconc for our Ministiy of Human Resources with the following
inquiry: wa$ il the casc that small businesses in Canada had ceased to be the creators of
employment in cyclical upswings because in so many cases staff wcre moving lowards
the status of hired consultants? ll is not an casy question to answer and the way we
altempled lo dcal with it was through our regisler of business. We Usted all Ihose OTall
firms which in tiie course of a period of three or four years (1990 to 1994) had lost
employment Wc divided them inlo Üiosc for whidb èie nominal (lhere was not much
inflatlon lo be conctancd widi) Qross Buísiness income (GBl) actually wenl up and thosc
for 'i^ch it weirt down. Nexl wc took those which had an increasii^ GBl and listed all
fijeif employees by tax number in the firsl year of the cxcrcise. We malched them against
their income tax statements in subsequent years and found how many had changed their
main source of income from wages and salaries lo income fixjm self-employcd. And
indeed, what our Ministry suspccted was the case.

I mentíon in dctúl this last exercise becausc I believe that labour market issues arc going
to be dominanl over the next decade pretty much cvcrywhere, 1 also believe lhal proper
answers can only be giv«i by extensive (but not exclusive) use of tax and other

8

adminislrative registcrs in addition to better classíilcalions of cmployment status and
some forni of longitudinal studics of labour market conditions.

Canyou trackyour laboutjlows in the way Idexcrihed ahove? In view of the way
modem technology and eeonomic organization are affecting the labour market, do
you thinkyou should develop the capaciíy to describe labour market flows
hngiíudinally? Have you consldered that you must gain access to administrative
registers of all klnds to be able to study labour dynamics (short of an extremely
expensive household xurvey) and that this is somethingfrom which ai presentyou do
not beneflt? Are your legal sysiems an impediment to access?

Are bytes more ímportant thaii dollars?

Not so long ago a new Canadian govcmment took over and announccd its social and
eeonomic agenda. The latter included a greal deal about the need to expand Canadian
markets abroad by an incrcasc in the competitive edge of Canadian exporting industry
and by marshalling Science and technology in order to aequire that competitive edge. ‘llie
party in power made great play of the fact that laige markeis mcan faster growth and
larger markets imply industrial innovaiion which in tum implies acquiríng faster and
more intelligenüy &e beneiils of seientifie discovery and technological applicatíon.

When it became necessary lo transiate these propositions into concrete policies the
Government realizcd that there was very liltie Information to suggest that any of this was
true. Moreover, if it was truc it was not at alJ obvious what should bc donc in order to,
for examplè. tighten the links between thosc who are at the foreíronl of scicntifíc
discovery and entrcprcncurs in charge of exporting enterprises; or how one should better
align one's supporl for seicncc and technology by spending on onc typc of research ratber
fhan anothci^ or vdiether it is pref<»able to support purc Science rather than technological
development or lhe othcr way round.

The fact of the matler is that all that Statisties Canada knew was how much govemment
and industry spent on something we defined and they called resscarch and developmentbut
there was no knowledge - not even conjectural •- of what were the outeomes of tliat
expenditure. We knew that the Oovemment, to starl at the obvious end, spent six or so
hillion dollars on a number of scientiiic pursuits and we had for each a veiy detailed
description of what it was. But no one in govemment could say whether similar results
could be atbúned by a ten per cent reduetion or increase in expenscs; or whether tlie
strueture of the expenditure was "optimal" in some crude sense.

The reason why wc knew this is was because the Federal Government - of all three
govemmental Icvcls in Canada - is the chief spender on scicnce. As is the case with any
other kind of federal expenditure, disbursements are distríbutcd by province and
provincial govemments are extremely keen to ensure, in the face of their eleclorales, that

9

thcj»^ have received at least thcir due, every time a fcdcrd expen^ture is lo bc shared
among provinces.

Accordingly, all lhal scicnce and technology statistics had been used for was to infoim
provincial electorates lhat thanks to the energies dcploycd by their respectivo
govemments» nol a penny that siiould accnic lo province A had been foolishly spcnt by
being granted to province B. Of course, if this is lhe chief purposc for which S&T
statistics are compiled, it is not welcomc to inlroduce another inquhy into thc purpose
and effectívmess of lhe expenditures on R&D.

As T write these Unes, we Iiave been given a credit to improve the quality of our
information on S&T and onc of thc fírst things we shall try and do is to see whether
tracking Information flows is at all possible. By this 1 mean that somewhcrc in the chain
of events that leads from having a good idea to converting it into a commercial success,
theie must be a series of steps by vdiich the original bytes of Information get channelled
to tíiose (receptors) who change it, adapt it and put it at the Service of others who in tum
convert it into something that is commercially usable.

Now if that chain of events invoives purchases and sales (for exampie. purchases of
patents or costs of reverse cnginccring) the chances are lhal we can track it But if no
change of financial assets is involved we arc not equipped to track the way in which bytes
of informaúon flow.

Surprisingly, at firet there was no great claim for ncw Information of the type I just
described. Indecd, it was thought that thc entire initiative could proceed with a minimum
of Information to steer it down the rígbt path. It was only at a laler stage that ofUciáls
reaiized that if thc govcmmcnt was to continue playing a role in lhe stimulation of
scientific research - and in the case of Cüanada just as in Bravdl govcmmcnt is involved in
big scientific initíatives - it had to become concemed with the availability of information
dtíferent iram thc kind of Information it had been used to.

The drive towards competltiveness and the requiremení to measure the role that R&D
plays in stimulating it are going to be with usfor the next while. However, only parí
of that role can be explained in lerms of money flows. Some (most?) of it is much
more meaningfuJ if accomted for in terms of "bytes". But none of us knows how to
detect, measure and compile such flows. llave you given any thought to the maíier?
Do you find it an interesting problem? Do you think that the demand that Canada
experiences today arui will continue to expeiience for the foreseeable future is likely
to hit Brazil if it has not done so already? And if it did (or when it does) whaí are you
going to answer?

10

Âre we uiorc productive Ihan wc look?

I hope so as citIzen bwt I have mixed fcciings as a statístician. Anyway, tfie problem is
very deep iirespcctíve of our hopes for a particular outcomes. As products (goods and
Services) become mcrcasingly more complex and tcchnology plays an mcrcasingly
important part in thcir dessign, thc issuc of whctiicr loday*s variant of a gívcn product is
qualitatively compaiablc to the variant thal was availabíe ten years ago has a veiy special
relevancc. Moreover, at no time in thc past was an equivalent conccra expressed largely
because it took longer to embody ncw tcchnologies in consumcr goods and Services and
so modify thcir quality dianicteristics.

Hiere are two propositions tbat are commonly expressed with varying degrees of
intensity:

• the quality of goods and sovices (including thc variety ofifered for choicc) is
inoreasing at a sleadíly íaster rate;

• statisticians measure ímperrcctly - or do not measure at all - thc variations in the
quality of lhe goods and Services thí^ price.

From these propositions it follows lhet wc also mis^measure our real output because so
much of it is estimated on the basis of dcflated values and the deflators stand indíeted of
being upwardly biased. Clearly, some important policy prescriptions follow ôom these
propositions. For example, targetíng lhe CPI annual per cent change at plus two per cent
maximum may in fact mean thal lhe largct is deflationtuy if the upvrard bias is of thc
order of two per cent or more. If it is felt that lhe economy ovcrhcats if it gro^vs at more
tiian four per <»nt annually, wc should target GüP growth not to exceed sáy two per cent
annuaUy because of tiie impKed downward bias involvcd in its estímation.

1 am not suggesling thal thc Canadian bias - or for that malter anyonc clsc*s - is of this
order of magnitude but rather that if these are lhe conscquaiccs of our findings it follows
toat we shall need a signifícanl ovohaul in thc manner in which we compile our pricc
indexes and in lhe manner in which we estimate and introduce our qualiQ^ adjustments. It
may seem esoleric for oíücc» which are struggling with the complexities of the new UN
System of nalional aceounts to bc wamcd that a few years down the line, thc growth rates
which thcy have so laboríously estimated will bc questioned and put up for subsUinlial
revision. My point is that they will not escape thc qucsüoning particularly if in thcir
country there is a sophisticated indushy opcn lo oulsidc influenoes and recording ever
fasler rates of adaptation to new ways of doing Ihings. Such, of course, is the ca.se of
Brazil which both for its gigantic internai market and for iis exporl performance cannot
kccp away from the latest in Chemicals, clcctronics and mwjhinery and equipmcnt

In the case of the United States, thc matter of persistent under-csslimatíon of real growth
has been broadied in the press and takcn up legislators in Con^ss. Thc CPI has been
sharply questioned and statístical agencies have led the way to table úie issue in
International venues largely to wam their peers that questioning of scemingly rock solld

11

numbers had staited. Wc in Canada, do not expect that we shall be immune from
questioning and wc havc started lo prepare oursclvcs for an intensive rcvicw of many of
our basic numbers so that we can at leas^ satisfy tlie inquiring public that we are oware of
lhe pFoblem and propose lo do what is warranted if possible in concert with our chief
parlncrs.

llow ahoutyou? Have echoes of these qmsHom heen heard? Âre yau asked to
revíew the matter of quality in íhe framework qfyour CPI? Are you concemed íhal
you may not be measuring gaim in output as well as you should? Do you think that
other biases - (here alt kinds ofbias in áll courúries ’ estimates - are of a naiure such
that they might offset thfs one? Have you considered whether you should go it alam
or do as we do - puí togeíher a group oflihe-minded colleagues and propose that we
labour share becausc the lask is too costly for our size and budget?

Conelusions

I do not know any more than you do what things will bc likc at the end of the decade.

But 1 am concemed that if wc makc no preparat‘«ons whatsocvor lo meet ^mt wc think is
likdy to happen, we shall look foolish and bc aceused of iack of professional
responsibüliy. For this reason, wherever wc csan wc takc anticipalory measures to at Icast
lay the groimdwork for initiatives wc believe we shall be asked to takc. Accoi^ingly» we
are preparing to have better Science and Technology statisties, we have started to aiign
colleagues to embark on a joint prico- quality adjujtment programmc, wc have resorted lo
a number of ambitious inquiries about the way in which the labour market works míd
what are the modes of labour r«nuneratíon, we have a study uiulerway on fomis of
industrial organization and we hope that by next ycar or at the latest, the year after we
diall be in a position to define a five to ten ycar programmc for the overhauling of our
econoraic stalislies. I believe that in doii^ so we are acting prudcntly. llicse are the
rcasons that led me to engage you in a discussion of why we should be rcasoning along
similar lines.